JPH0993547A - Motion detection circuit - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect a motion at an edge part of an image more accurately. SOLUTION: A high pass filter 14 extracts a high frequency component of a current frame signal S0 and a high pass filter 21 extracts a high frequency component of a signal S4 of two-preceding frame. A subtractor 12 and an absolute value circuit 13 take an absolute value of the signals S0, S4 to obtain an inter-2-frame motion detection signal 604. A logic circuit 22 obtains a control signal E04 by outputs of the filters 14,21. A sensitivity control circuit 16 decreases the inter-2-frame motion detection signal Δ04 when the control signal E04 is larger and increases the inter-2-frame motion detection signal Δ04 when the control signal E04 is smaller.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention moves an image of an image signal (sub-sample transmission signal) which is sub-sampled and encoded at an n frame period (2 frame period for a MUSE signal) like a MUSE decoder. In a device that adaptively switches interpolation processing between a portion and a still portion, the present invention relates to a motion detection circuit for determining a moving portion and a still portion of an input image signal, , A motion detection circuit adapted to adaptively control the sensitivity of motion detection according to an image pattern.

[0002]

2. Description of the Related Art In recent years, high-definition television (HDTV) broadcasting by the MUSE system has been in full swing.
In the MUSE method, offset sub-sampling (decimation) is performed on the encoding side between fields and frames. Therefore, in order to reproduce a still image on the decoding side, it is necessary to combine (interpolate between frames and between fields) from MUSE signals for four fields. However, if this still image processing is applied to a moving part, a multiple image is formed. Therefore, a moving part is detected from the MUSE signal, and an image is constructed from the signal for one field in that part (within the field). Insert). Therefore, the MUSE decoder needs a process (motion detection process) for discriminating between a moving image part and a still part.

Here, the schematic structure and operation of the MUSE decoder will be described with reference to FIG. In FIG.
The MUSE signal S0 input from the input terminal 1 is input to the motion detection circuit 2, the moving image processing circuit 3, and the still image processing circuit 4. The moving image processing circuit 3 synthesizes an image from signals of one field (field interpolation), and the still image processing circuit 4 synthesizes images from signals of four fields (interframe interpolation and interfield interpolation). The motion detection circuit 2 detects a motion part of the image, and controls the mixing circuit 5 with this motion detection signal. The mixing circuit 5 mixes the signal subjected to the moving image processing by the moving image processing circuit 3 and the signal subjected to the still image processing by the still image processing circuit 4 at a ratio according to the amount of motion (motion detection signal) and outputs the mixed signal.

When the output of the motion detection circuit 2 is k, the output of the moving image processing circuit 3 is Sm, and the output of the still image processing circuit 4 is Ss, the output y of the mixing circuit 5 is given by the following equation. y = k × Sm + (1−k) × Ss However, k satisfies the condition of 0 ≦ k ≦ 1, and is 0 in the case of a complete still image and 1 in the case of a complete moving image. Detailed description of the moving image processing circuit 3, the still image processing circuit 4, and the mixing circuit 5 will be omitted.

Next, the configuration and operation of the conventional motion detection circuit 2 shown in FIG. 4 will be described with reference to FIGS. 5 and 6. By the way, it is preferable that the motion detection circuit 2 detects the motion amount based on the difference for each pixel in one frame.
In the E signal, since the sampling pattern makes one round in four fields (two frames), the same sample points do not exist in one frame, and it is difficult to detect motion. Therefore, in order to perform accurate motion detection, it is necessary to use the difference between two frames whose sample points match.

In FIG. 5, the MUSE signal S0 received from the input terminal 10 is a signal S4 delayed by two frames by the two-frame delay memory 11. The subtractor 12 obtains the difference between the signal S0 of the current frame and the signal S4 of the previous frame. An absolute value circuit (ABS) 13 converts the output of the subtractor 12 into an absolute value and obtains a 2-frame motion detection signal Δ04. On the other hand, the signal S0 is the high-pass filter (HPF) 1
4 is also input, and the high frequency component is extracted here. After that, the high frequency signal is converted into an absolute value by the absolute value circuit (ABS) 15, and the edge detection signal E0 is obtained. The edge detection signal E0 is input to the sensitivity control circuit 16 together with the 2-frame motion detection signal Δ04, and the sensitivity of the 2-frame motion detection signal Δ04 is controlled according to the edge detection signal E0. After that, the field memories 17, 18, 19 and the maximum value extraction circuit 20 perform temporal expansion (fill-in) to obtain a final motion detection signal k.

The reason why the edge detection signal E0 controls the sensitivity of the two-frame motion detection signal Δ04 will be described. In FIG. 6, consider a motion detection operation when two types of objects A and B slightly move. The object A is a case where the edge of the image is large, and the object B is a case where the edge is small. Even though the amplitudes of the two objects are almost equal, the two-frame motion detection signal Δ04 causes large motion detection only in the edge portion of the object A. In this case, for example, even if there is jitter in the sampling clock or slight camera shake during camera shooting, large motion detection is unfavorable. Therefore, the edge portion is controlled so as to suppress the detection sensitivity according to the magnitude of the edge detection signal E0 which is the magnitude of the edge. Incidentally, as an example corresponding to this conventional example, Japanese Patent Laid-Open No. 62-17
There is No. 2877.

[0008]

However, according to the above-described control method of the motion detection sensitivity in the conventional motion detection circuit, the sensitivity is uniformly lowered in a portion having a large edge regardless of the degree of the motion, and The motion detection is likely to be insufficient for a portion where the image is intense, resulting in a problem of multiple images in that portion. On the other hand, if the sensitivity control for the edge portion is relaxed, the edge portion is processed as a moving image when the movement is minute such as jitter or camera shake, which is not preferable. Originally, it is not necessary to lower the sensitivity for a part that is moving violently even if the edge is large, and video processing should be performed, on the contrary, it is extremely sensitive to minute movements such as jitter and camera shake. Should be lowered to process still images. Therefore, if the sensitivity of motion detection is controlled only by the size of the edge as in the conventional method, the only solution is to set an appropriate compromise point, and accurate motion detection cannot be expected. Especially,
More accurate because the difference in image quality when making an erroneous judgment is extremely different in the image of the edge part (multiple image failure when processing a still image of a moving image or aliasing interference when processing a moving image of a still image becomes noticeable) Motion detection is required.

The present invention has been made in view of the above problems, and an object of the present invention is to provide a motion detection circuit capable of more accurately detecting motion at an edge portion of an image.

[0010]

SUMMARY OF THE INVENTION The present invention is a motion detection circuit used in a device for decoding a sub-sampled transmission signal in order to solve the above-mentioned problems of the prior art. n is an integer equal to or greater than 1) delay means (11) for delaying the frame, and n frames by the absolute value of the difference between the input image signal (S0) and the n frame delay signal (S4) obtained by the delay means. Inter-frame motion detection means (1) for obtaining inter-motion detection signal (Δ04)
2, 13), a first high-pass filter (14) for extracting a high-pass component (H0) of the input image signal, and a high-pass component (n) of the n-frame delay signal (2). A second high-pass filter (21) for extracting H4), a high-pass component of the input image signal extracted by the first high-pass filter, and the second high-pass filter Provided is a motion detecting circuit comprising level changing means (16, 22) for changing the level of the motion detection signal between the n frames according to the high frequency component of the n frame delay signal thus generated. To do.

[0011]

BEST MODE FOR CARRYING OUT THE INVENTION A motion detection circuit of the present invention will be described below with reference to the accompanying drawings. FIG. 1 is a block diagram showing an embodiment of a motion detecting circuit of the present invention, FIG. 2 is a block diagram showing a concrete configuration of a logic circuit in FIG. 1, and FIG. 3 is a diagram for explaining the operation of the motion detecting circuit of the present invention. FIG. In FIG. 1, the same parts as those in FIG. 5 are designated by the same reference numerals.

In FIG. 1, the MUSE signal S0 input from the input terminal 10 is a signal S4 delayed by two frames by the two-frame delay memory 11. The subtractor 12 obtains the difference between the signal S0 of the current frame and the signal S4 of the previous frame. An absolute value circuit (ABS) 13 converts the output of the subtractor 12 into an absolute value and obtains a 2-frame motion detection signal Δ04. Signal S output from 2-frame delay memory 11
4 is also input to a high pass filter (HPF) 21, where high frequency components are extracted and input to a logic circuit 22.

On the other hand, the signal S0 is a high-pass filter (HP
F) 14 is also input, and high frequency components are extracted here.
This high frequency signal is input to the logic circuit 22. The logic circuit 22 generates a control signal E04 for controlling the sensitivity of motion detection by the sensitivity control circuit 16 from the outputs of the HPFs 14 and 21 as described later. This control signal E04
Is input to the sensitivity control circuit 16 together with the inter-two-frame motion detection signal Δ04, and the inter-two-frame motion detection signal Δ04 has its sensitivity controlled according to the control signal E04. After that, the field memories 17, 18, 19 and the maximum value extraction circuit 20 perform temporal expansion (fill-in) to obtain a final motion detection signal k.

The logic circuit 22 will now be described. The logic circuit 22 is a circuit for generating a control signal E04 for controlling the sensitivity from both the output signal H0 of the HPF 14 and the output signal H4 of the HPF 21 based on the pattern difference and the edge size. This is to detect the degree of movement, whether it is a small movement such as jitter or camera shake, or a movement that changes drastically. Since the HPF 14 and the HPF 21 here are mainly for detecting the gradient of the waveform, a simple differential filter is sufficient.

As an example, the logic circuit 22 is constructed as shown in FIG. This logic circuit 22 receives the signal H
The absolute value of the difference between 0 and the signal H4 is obtained, and the amount of edge is controlled according to this magnitude. That is, in FIG. 2, the signal H0
And the signal H4 are input to the subtraction circuit 221, and the signal H4 is subtracted from the signal H0. The output of the subtraction circuit 221 is converted into an absolute value by the absolute value circuit 223. This output is Δ
04H. On the other hand, the signals H0 and H4 are converted into absolute values by the absolute value circuits 222 and 224, respectively. As a result, the edge amounts of the signal S0 of the current frame and the signal S4 two frames before are obtained. Absolute value circuits 222, 224
Is output to the maximum value extraction circuit 225, and the larger value of these two is selected and output. This output is H04. In this embodiment, the edge amount of the image signal is obtained by both the signal S0 of the current frame and the signal S4 of two frames before, but either one may be used. In this case, the maximum value extraction circuit 225 can be omitted.

The signal Δ0 output from the absolute value circuit 223
4H and the signal H04 output from the maximum value extraction circuit 225
Is input to the sensitivity control circuit 226. Sensitivity control circuit 22
6 controls the signal H04 by the signal Δ04H. That is, the sensitivity control circuit 226 performs sensitivity control so as to lower the level of the signal H04 when the signal Δ04H is large, and conversely, raises the level of the signal H04 (or does not simply lower the level) when the signal Δ04H is small. To do. As a result, when the signal Δ04H is large, the control signal E04 is small, and conversely, when the signal Δ04H is small, the control signal E0 is small.
4 grows (or does not shrink). The output of the sensitivity control circuit 226 is input to the sensitivity control circuit 16 in FIG. 1 as the control signal E04. Then, the sensitivity control circuit 16
When the input control signal E04 is large, the level of the inter-two-frame motion detection signal Δ04 is lowered to reduce the sensitivity of motion detection, and when it is small, the inter-two-frame motion detection signal Δ0.
Increase the level of 4 to increase the sensitivity of motion detection.
Output as

The state of these series of processes will be described with reference to FIG. In FIG. 3, the signal S0 in FIG.
S4, Δ04, H0, H4, Δ04H, H04, E0
4, K04 is shown. In FIG. 3, the object A has a small movement in the edge portion, and the object C shows a case in which the edge portion is moving quickly. When the movement of the edge part is small like the object A, the difference (Δ04H) between the waveforms of the signal H0 and the signal H4 is small, but when the object C is moving quickly, the signal H0 and the signal H0 are different from each other. The difference (Δ04H) in the waveform of H4 is large. Although these two objects A and C are almost equal in magnitude with respect to the signal H04, the signal Δ
The magnitude of the signal E04 controlled by 04H is obviously different. In this example, the signal K04 after the sensitivity control is small for the object A having a small motion and large for the object C having a large motion.

As described above, according to the present invention, unlike the conventional motion detection sensitivity control method in the motion detection circuit which uniformly controls the sensitivity of motion detection only by the size of the edge, the difference in the edge waveforms causes the motion detection. By contributing to the sensitivity control of, the degree of motion can be reflected, and as a result, more accurate motion detection becomes possible. Further, it is possible to alleviate the problem that a still image is detected in motion at the time of fade-in and fade-out. Note that the configurations of FIGS. 1 and 2 are merely an embodiment of the present invention, and the difference between the high frequency signal of the current frame and the signal of n frames past of this signal may be used for controlling the motion detection sensitivity. However, it is not limited to these configurations.

From the above, in the motion detection circuit of the present invention,
The motion detection in the edge portion of the image can be performed more accurately. Since fold-back interference that occurs when processing minute movements in moving images and multiple image failure when processing still images of strongly moving parts are prominent at the edge part of the image,
The effect of the sensitivity control method according to the present invention, which utilizes the change in the pattern of the edge portion, is particularly effective. According to the configuration shown in FIG. 1, the increase in the number of circuits from the conventional configuration is very small, and the cost does not increase in the recent IC integration technology.

[0020]

As described in detail above, the motion detection circuit of the present invention includes the first high pass filter for extracting the high frequency component of the input image signal and the high frequency component of the n-frame delay signal. A high-pass component of the input image signal extracted by the first high-pass filter and a second high-pass filter for extracting
Since the configuration is provided with the level varying means for varying the level of the motion detection signal for n frames depending on the high frequency component of the frame delay signal, whether the image is moving violently or whether it is a minute movement due to the influence of jitter or the like. The degree of motion can be detected, and more accurate sensitivity control of motion detection sensitivity can be performed using this detection result.

[Brief description of drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention.

FIG. 2 is a block diagram showing a specific configuration of a logic circuit 22 in FIG.

FIG. 3 is a waveform chart for explaining the operation of the present invention.

FIG. 4 is a block diagram showing a schematic configuration of a MUSE decoder.

FIG. 5 is a block diagram showing a conventional example.

FIG. 6 is a waveform diagram for explaining the operation of the conventional example.

[Explanation of symbols]

 11 2 Frame Delay Memory (Delay Means) 12 Subtractor (n-frame Motion Detection Means) 13 Absolute Value Circuit (n-frame Motion Detection Means) 14 High-Pass Filter (First High-Pass Filter) 16 Sensitivity Control Circuit 17- 19 field memory 20 maximum value extraction circuit 21 high-pass filter (second high-pass filter) 22 logic circuit (control signal generation circuit)

Claims (3)

[Claims] 1. A motion detection circuit used in an apparatus for decoding a sub-sample transmission signal, comprising: delay means for delaying an input image signal by n (n is an integer of 1 or more) frames; and the input image signal. And an n-frame motion detection means for obtaining an n-frame motion detection signal based on the absolute value of the difference between the n-frame delay signal obtained by the delay means and the n-frame delay detection signal. A first high-pass filter for extracting a component, a second high-pass filter for extracting a high-pass component of the n-frame delayed signal, and the input extracted by the first high-pass filter The level of the motion detection signal between the n frames can be adjusted by the high frequency component of the image signal and the high frequency component of the n frame delay signal extracted by the second high pass filter. Motion detection circuit, characterized in that constructed by a level varying means for. 2. The level varying means, when the absolute value of the difference between the output of the first high-pass filter and the output of the second high-pass filter is small, the level of the motion detection signal between n frames. And the level of the motion detection signal between n frames is lowered when the absolute value of the difference between the output of the first high pass filter and the output of the second high pass filter is large. Item 1. The motion detection circuit according to item 1. 3. The level varying means for controlling the sensitivity control circuit by a sensitivity control circuit for controlling the sensitivity of the motion detection signal between n frames and the outputs of the first and second high pass filters. And a control signal generating circuit for generating a control signal of the control signal generating circuit, wherein the control signal generating circuit calculates an absolute value of a difference between an output of the first high-pass filter and an output of the second high-pass filter. First means for obtaining the edge amount, second means for obtaining the edge amount of the image signal by at least one of the output of the first high-pass filter and the output of the second high-pass filter, and the first means. 3. The motion detecting circuit according to claim 1, further comprising: a third means for reducing the level of the output of the second means as the output of the means increases.